Extracorporeal medical machines such as the pump-oxygenator and artificial kidney rely on systemic heparinization to improve blood compatibility. However, heparin can lead to serious complications such as bleeding. With the prospect of even longer perfusion times with machines such as the membrane oxygenator, the problems due to heparinization become more severe. Although many approaches have been explored to solve this problem, such as the use of neutralizing compounds or heparin bonded surfaces, there still remains no real alternative to systemic heparinization. We propose a new method to control heparin levels using a blood filter containing immobilized heparinase. Such a filter might be used in situations where it is desired to heparinise the extracorporeal circuit without simultaneous heparinization of the patient. Alternatively, it could eliminate the use of neutraliziang compounds such as protamine. Because the amount of data on heparinase has, until now, been limited and the methods of producing it are inadequate for large scale use, research has focused not only on the development and testing of the filter but on enzyme production and purification. Thus far we have been able (1) to increase volumetric enzyme production over a thousand fold from previous published procedures, (2) to purify heparinase by over 1000 fold; the heparinase is now electrophoretically pure; (3) to characterize the biochemical properties of heparinase and isolate the first heparinase inhibitor; (4) to immobilize heparinase with 91 percent activity recovery and excellent stability; (5) to design a filter that is capable of degrading over 99 percent of heparin's anticoagulant activity in 2-6 minutes in human blood in vitro and in canine blood in vivo. Having made these findings, the objectives of the current study are (1) to explore the use of genetic engineering to develop a method to produce large quantities of heparinase without the other impurities commonly associated with current procedures; (2) to determine if there is a metabolic build-up of heparin degradation products and to use other toxicological tests to explore possible toxicity; 3) To characterize the immobilized enzyme system with respect to kinetic parameters and optimally design a small reactor capable of degrading a clinically used amount of heparin at flow rates of up to 250 ml/min. and 4) to test the immobilized enzyme reactor in a sheep hemodialysis model for efficacy and safety.